Systems and methods for filtering microblogs

Abstract

Systems, devices and methods are provided for filtering microblogs. For example, one or more first subject terms are acquired based on at least information associated with content of a first microblog to be filtered; an attention term list is acquired corresponding to an user identification; one or more first degrees of attention corresponding to the first subject terms are acquired based on at least information associated with the attention term list; and whether to filter out the first microblog is determined based on at least information associated with the first degrees of attention.

The invention claimed is: 
     
       1. A method for filtering microblogs, the method comprising:
 acquiring one or more first subject terms based on at least information associated with content of a first microblog to be filtered; 
 acquiring an attention term list corresponding to a user identification, the attention term list comprising a plurality of subject terms, and each subject term in the attention term list corresponding to a degree of attention which indicates an extent of user attention on a subject of the subject term, wherein the degree of attention of each subject term in the attention term list is determined based on at least one of preference settings or microblog interaction history of a user having the user identification; 
 acquiring one or more first degrees of attention corresponding to the first subject terms based on at least information associated with the attention term list; 
 determining whether to filter out the first microblog based on at least information associated with the first degrees of attention; 
 in response to determining not to filter out the first microblog, pushing the first microblog to a user terminal logged in with the user identification, such that the user terminal receives the first microblog containing the content fitting user interest; and 
 in response to determining to filter out the first microblog, excluding the first microblog from information pushed to the user terminal, such that dissemination of junk information is avoided and network resources are saved, 
 wherein determining whether to filter out the first microblog comprises: 
 obtaining total degrees of attention corresponding to the one or more first subject terms, including: for each of the one or more the first subject term,
 detecting one or more related subject terms of a first subject term based on a preset mapping table, each of the one or more related subject terms corresponding to a second degree of attention which indicates an extent of attention of the user having the user identification; 
 acquiring one or more degrees of relation corresponding to the detected one or more related subject terms based on the preset mapping table, each degree of relation describing a relevance degree between a related subject term and the first subject term; and 
 calculating a total degree of attention corresponding to the first subject term by combining a first degree of attention of the user corresponding to the first subject term, the one or more degrees of relation corresponding to the detected one or more related subject terms, and one or more second degrees of attention corresponding to the one or more related subject terms; and 
 
 determining whether to filter out the first microblog by comparing the total degrees of attention with a filtering threshold. 
 
     
     
       2. The method of  claim 1 , wherein the acquiring one or more first subject terms based on at least information associated with content of a first microblog includes:
 performing word segmentation for the first microblog to acquire one or more first candidate terms; and 
 searching for the first candidate terms in a keyword dictionary to be the first subject terms. 
 
     
     
       3. The method of  claim 1 , wherein the acquiring one or more first degrees of attention corresponding to the first subject terms based on at least information associated with the attention term list includes:
 searching for the first subject terms in the attention term list; 
 acquiring the first degrees of attention corresponding to the first subject terms in the attention term list; and 
 determining whether to filter out the first microblog based on at least information associated with the first degrees of attention corresponding to the first subject terms in the attention term list. 
 
     
     
       4. The method of  claim 3 , the method further comprising:
 acquiring a sample set; 
 detecting a first number of times related to occurrence of one or more third terms of a keyword dictionary in the sample set and a second number of times related to co-occurrence of fourth pairwise terms in the sample set; and 
 calculating one or more third degrees of relation related to the fourth pairwise terms based on at least information associated with the first number of times and the second number of times. 
 
     
     
       5. The method of  claim 1 , wherein the acquiring one or more first degrees of attention corresponding to the first subject terms based on at least information associated with the attention term list and the determining whether to filter out the first microblog based on at least information associated with the first degrees of attention include:
 acquiring one or more second terms related to the first subject terms, the second terms and the first subject terms being associated with one or more second degrees of relation; 
 searching for the second terms in the attention term list; 
 acquiring one or more second degrees of attention corresponding to the second terms in the attention term list and the second degrees of relation; and 
 determining whether to filter out the first microblog based on at least information associated with the second degrees of attention and the second degrees of relation. 
 
     
     
       6. The method of  claim 1 , further comprising:
 acquiring an input blacklist and an input whitelist; 
 determining whether the first subject terms exist in the whitelist; 
 in response to the first subject terms existing in the whitelist, determining not to filter out the first microblog; and 
 in response to the first subject terms not existing in the whitelist, 
 determining whether the first subject terms exist in the blacklist; and 
 in response to the first subject terms existing in the blacklist, determining to filter out the first microblog. 
 
     
     
       7. The method of  claim 1 , further comprising:
 adding one or more fifth terms in the blacklist and one or more sixth terms in the whitelist to the attention term list; and 
 setting one or more third degrees of attention corresponding to the fifth terms and the sixth terms. 
 
     
     
       8. The method of  claim 1 , further comprising:
 in response to the first microblog not being filtered out, acquiring feedback information of the first microblog; and 
 adjusting the first degrees of attention corresponding to the first subject terms in the attention term list based on at least information associated with the feedback information. 
 
     
     
       9. The method of  claim 1 , further comprising:
 acquiring a second microblog corresponding to the user identification; 
 performing word segmentation for the second microblog to acquire one or more second candidate terms; 
 searching for the second candidate terms in a keyword dictionary; 
 acquiring a word frequency of the second candidate terms; 
 calculating one or more fourth degrees of attention related to the second candidate terms based on at least information associated with the word frequency; and 
 generating the attention term list corresponding to the user identification based on at least information associated with the fourth degrees of attention. 
 
     
     
       10. The method of  claim 1 , wherein acquiring one or more degrees of relation corresponding to the detected one or more related subject terms comprises:
 acquiring a sample set including a plurality of microblogs, wherein a plurality of subject terms occur in the sample set; 
 detecting a first number of occurrences of the first subject term in the sample set; and 
 for each of the one or more related subject term of the first subject term,
 detecting a second number of occurrences of a related subject term in the sample set; 
 detecting a third number of co-occurrences of the first subject term and the related subject term in the sample set; and 
 calculating a degree of relation corresponding to the related subject term based on the first number, the second number, and the third number. 
 
 
     
     
       11. A device for filtering microblogs, the device comprising:
 a memory; and 
 a processor coupled to the memory and configured to: 
 acquire one or more first subject terms based on at least information associated with content of a first microblog to be filtered; 
 acquire an attention term list corresponding to a user identification, the attention term list comprising a plurality of subject terms, and each subject term in the attention term list corresponding to a degree of attention which indicates an extent of user attention on a subject of the subject term, wherein the degree of attention of each subject term in the attention term list is determined based on at least one of preference settings or microblog interaction history of a user having the user identification; and 
 acquire one or more first degrees of attention corresponding to the first subject terms based on at least information associated with the attention term list; 
 determine whether to filter out the first microblog based on at least information associated with the first degrees of attention; 
 in response to determining not to filter out the first microblog, push the first microblog to a user terminal logged in with the user identification, such that the user terminal receives the first microblog containing the content fitting user interest; and 
 in response to determining to filter out the first microblog, exclude the first microblog from information pushed to the user terminal, such that dissemination of junk information is avoided and network resources are saved, 
 wherein when determining whether to filter out the first microblog, the processor is further configured to: 
 obtain total degrees of attention corresponding to the one or more first subject terms, including: for each of the one or more the first subject term,
 detect one or more related subject terms of a first subject term based on a preset mapping table, each of the one or more related subject terms corresponding to a second degree of attention which indicates an extent of attention of the user having the user identification; 
 acquire one or more degrees of relation corresponding to the detected one or more related subject terms based on the preset mapping table, each degree of relation describing a relevance degree between a related subject term and the first subject term; and 
 calculate a total degree of attention corresponding to the first subject term by combining a first degree of attention of the user corresponding to the first subject term, the one or more degrees of relation corresponding to the detected one or more related subject terms, and one or more second degrees of attention corresponding to the one or more related subject terms; and 
 
 determine whether to filter out the first microblog by comparing the total degrees of attention with a filtering threshold. 
 
     
     
       12. The device of  claim 11 , wherein the processor is further configured to:
 perform word segmentation for the first microblog to acquire one or more first candidate terms; and 
 search for the first candidate terms in a keyword dictionary to be the first subject terms. 
 
     
     
       13. The device of  claim 11 , wherein the processor is further configured to:
 search for the first subject terms in the attention term list; 
 acquire the first degrees of attention corresponding to the first subject terms in the attention term list; and 
 determine whether to filter out the first microblog based on at least information associated with the first degrees of attention corresponding to the first subject terms in the attention term list. 
 
     
     
       14. The device of  claim 13 , wherein the processor is further configured to:
 acquire a sample set; 
 detect a first number of times related to occurrence of one or more third terms of a keyword dictionary in the sample set and a second number of times related to co-occurrence of fourth pairwise terms in the sample set; and 
 calculate one or more third degrees of relation related to the fourth pairwise terms based on at least information associated with the first number of times and the second number of times. 
 
     
     
       15. The device of  claim 11 , wherein the processor is further configured to:
 acquire one or more second terms related to the first subject terms, the second terms and the first subject terms being associated with one or more second degrees of relation and search for the second terms in the attention term list; and 
 acquire one or more second degrees of attention corresponding to the second terms in the attention term list and the second degrees of relation and determine whether to filter out the first microblog based on at least information associated with the second degrees of attention and the second degrees of relation. 
 
     
     
       16. The device of  claim 11 , wherein the processor is further configured to:
 acquire an input blacklist and an input whitelist; 
 determine whether the first subject terms exist in the whitelist and, in response to the first subject terms existing in the whitelist, determine not to filter out the first microblog; 
 in response to the first subject terms not existing in the whitelist, determine whether the first subject terms exist in the blacklist and, in response to the first subject terms existing in the blacklist, determine to filter out the first microblog. 
 
     
     
       17. The device of  claim 11 , wherein the processor is further configured to:
 in response to the first microblog not being filtered out, acquire feedback information of the first microblog and adjust the first degrees of attention corresponding to the first subject terms in the attention term list based on at least information associated with the feedback information. 
 
     
     
       18. The device of  claim 11 , wherein the processor is further configured to:
 acquire a second microblog corresponding to the user identification and perform word segmentation for the second microblog to acquire one or more second candidate terms; 
 search for the second candidate terms in a keyword dictionary; 
 acquire a word frequency of the second candidate terms and calculate one or more fourth degrees of attention related to the second candidate terms based on at least information associated with the word frequency; and 
 generate the attention term list corresponding to the user identification based on at least information associated with the fourth degrees of attention. 
 
     
     
       19. The device of  claim 11 , wherein when acquiring the one or more degrees of relation corresponding to the detected one or more related subject terms, the processor is further configured to:
 acquire a sample set including a plurality of microblogs, wherein a plurality of subject terms occur in the sample set; 
 detect a first number of occurrences of the first subject term in the sample set; and 
 for each of the one or more related subject term of the first subject term,
 detect a second number of occurrences of a related subject term in the sample set; 
 detect a third number of co-occurrences of the first subject term and the related subject term in the sample set; and 
 calculate a degree of relation corresponding to the related subject term based on the first number, the second number, and the third number. 
 
 
     
     
       20. A non-transitory computer readable storage medium comprising programming instructions for filtering microblogs, the programming instructions configured to cause one or more data processors to execute operations comprising:
 acquiring one or more first subject terms based on at least information associated with content of a first microblog to be filtered; 
 acquiring an attention term list corresponding to a user identification, the attention term list comprising a plurality of subject terms, and each subject term in the attention term list corresponding to a degree of attention which indicates an extent of user attention on a subject of the subject term, wherein the degree of attention of each subject term in the attention term list is determined based on at least one of preference settings or microblog interaction history of a user having the user identification; 
 acquiring one or more first degrees of attention corresponding to the first subject terms based on at least information associated with the attention term list; 
 determining whether to filter out the first microblog based on at least information associated with the first degrees of attention; 
 in response to determining not to filter out the first microblog, pushing the first microblog to a user terminal logged in with the user identification, such that the user terminal receives the first microblog containing the content fitting user interest; and 
 in response to determining to filter out the first microblog, excluding the first microblog from information pushed to the user terminal, such that dissemination of junk information is avoided and network resources are saved, 
 wherein determining whether to filter out the first microblog comprises: 
 obtaining total degrees of attention corresponding to the one or more first subject terms, including: for each of the one or more the first subject term,
 detecting one or more related subject terms of a first subject term based on a preset mapping table, each of the one or more related subject terms corresponding to a second degree of attention which indicates an extent of attention of the user having the user identification; 
 acquiring one or more degrees of relation corresponding to the detected one or more related subject terms based on the preset mapping table, each degree of relation describing a relevance degree between a related subject term and the first subject term; and 
 calculating a total degree of attention corresponding to the first subject term by combining a first degree of attention of the user corresponding to the first subject term, the one or more degrees of relation corresponding to the detected one or more related subject terms, and one or more second degrees of attention corresponding to the one or more related subject terms; and 
 
 determining whether to filter out the first microblog by comparing the total degrees of attention with a filtering threshold. 
 
     
     
       21. The storage medium of  claim 20 , wherein acquiring one or more degrees of relation corresponding to the detected one or more related subject terms comprises:
 acquiring a sample set including a plurality of microblogs, wherein a plurality of subject terms occur in the sample set; 
 detecting a first number of occurrences of the first subject term in the sample set; and 
 for each of the one or more related subject term of the first subject term,
 detecting a second number of occurrences of a related subject term in the sample set; 
 detecting a third number of co-occurrences of the first subject term and the related subject term in the sample set; and 
 calculating a degree of relation corresponding to the related subject term based on the first number, the second number, and the third number.